Harmful algal blooms (HABs) are more than seasonal nuisances- they are escalating threats to ecosystems, economies, and public health.

Traditional monitoring often lags behind the rapid spread of HABs, leaving responders one step behind.

Our Approach: Precision Monitoring & Forecasting
Our AI models extensively use satellite remote sensing and in-situ environmental measurements, integrating these into scalable, high-frequency tools for ecosystem monitoring.

• They map bloom activity across lakes, with a monthly rolling forecast for the upcoming 12 months, producing value-added insights on the movement, extent, and risk factors for cyanobacterial blooms, which enable early, informed action.

• We have also developed an explainability layer that provides insight into bloom causation, supporting targeted prevention.

• The models have been designed with scalability in mind. Once calibrated using local in-situ data, they can operate autonomously, making them ideal for deployment in remote or under-resourced regions.

Tried and Tested
Our models were rigorously tested on the HABs at Lough Neagh (Northern Ireland). The forecasting model achieved a mean absolute percentage error (MAPE) of 27.8% for chlorophyll concentration predictions, translating to a general predictive accuracy within ±30%.